Process Forecasting: Towards Proactive Business Process Management

Abstract

The digital economy is highly volatile and uncertain. Ever-changing customer needs and technical progress increase the pressure on organizations to continuously improve and innovate their business processes. The ability to anticipate incremental and radical process changes required in the future is a critical success factor. However, organizations often fail to forecast future business process designs and process performance. One reason is that Business Process Management (BPM) is dominated by reactive methods (e.g., lean management, traditional process monitoring), whereas there are only a few future-oriented approaches (e.g., process simulation, predictive process monitoring). This paper supports the shift towards proactive BPM by coining the notion of process forecasting-an umbrella concept for future-oriented BPM methods and techniques. We motivate the need for process forecasting by eliciting various types of process forecasting from BPM use cases and create a first understanding of its scope by providing a definition, a reference process, showing the steps to be followed in process forecasting initiatives, and a positioning against related BPM sub-areas. The definition and reference process are based on a structured literature review.

Building a complementary agenda for business process management and digital innovation

“…This problem is commonplace during "as-is" process modellingoften a considerable time and cost investment with unclear, indirect, or indeterminate benefits . Moreover, by limiting BPM to existing problems, managers forego opportunities for solutions that involve proactive approaches, such as searching and implementing solutions before problem arises (Benner & Tushman, 2003;Poll et al, 2018). If BPM is purely reactive, managers would need to wait for digital innovations to impact existing business processes and manifest problems that require "fixing".…”

Process Model Forecasting Using Time Series Analysis of Event Sequence Data

“…To this end, we complement the model-level prediction technique with a visualisation system to enable analysts to understand the forthcoming changes to the processes. Various process analysis tasks benefit from process forecasting [20], most notably process forecasting helps understanding the incremental changes and adaptations that happen to the process model and to project them into the future. In terms of visualisation principles, we follow the "Visual Information-Seeking Mantra": overview first, zoom and filter, then details-ondemand [25].…”

“…Many analytical tasks require not only an understanding of the current as-is, but also the anticipated will-be process model. A key challenge in this context is the consideration of evolution as processes are known to be subject to drift [15,20,31,32]. A forecast can then inform the process analyst how the will-be process model might differ from the current as-is if no measures are taken, e.g., against emerging bottlenecks.…”

Encoder-Decoder Generative Adversarial Nets for Suffix Generation and Remaining Time Prediction of Business Process Models

“…First, we show that our framework systematically outperforms the baselines in terms of accuracy, despite only using the order and completion time of events as features, as opposed to some of the baselines, which also use other available attributes in the log, such as the resource associated with each process activity. Second, our framework provides the k most probable suffixes for a prefix that can be used for further analysis such as process forecasting [24].…”